The Orontids of Armenia by Cyril Toumanoff This study appears as part III of Toumanoff's Studies in Christian Caucasian History (Georgetown, 1963), pp. 277-354. An earlier version appeared in the journal Le Museon 72(1959), pp. 1-36 and 73(1960), pp. 73-106. The Orontids of Armenia Biblio graphy, pp. 501-523 Maps appear as an attachment to the present document. This material is presented solely for non-commercial educational/research purposes. I 1. The genesis of the Armenian nation han been examined in an earlier Study.* Us nucleus, succeeding to the role of the Vannic nucleus ot Urarfu, was the ‘proto-Armenian,’ Hayasa-Phrygian, people-state,® which at first oc¬ cupied only a small section of the former Urartian, or subsequent Armenian, territory. And it was, precisely, of the expansion of this people-state over that territory, and of its blending witli the remaining Urartians and other proto- Caucasians that the Armenian nation was bom. That expansion proceeded from the earliest proto-Armenian settlement in the basin of the Arsanias (East¬ ern Euphrates) up the Euphrates, to the valley of the upper Tigris, and espe¬ cially to that of the Araxes, which is the central Armenian plain,® This expand¬ ing proto-Armenian nucleus formed a separate satrapy in the Iranian empire^ while the rest of the inhabitants of the Armenian Plateau, both the remaining Urartians and other proto-Caucasians, were included in several other satrapies.* Between Herodotus’s day and the year 401, when the Ten Thousand passed through it, the land of the proto-Armenians had become so enlarged as to form, in addition to the Satrapy of Armenia, also the trans-Euphratensian vice-Sa- trapy of West Armenia.^ This division subsisted in the Hellenistic phase, as that between Greater Armenia and Lesser Armenia, From Xenophon we gather a few glimpses of the social order that existed under satrapal control in cis-Caucasia: of the proto-Armenian clan-heads, ruling fortified rural settlements {xo>(jtdgxv^> dQxv fieyd^ov zov >tai ’Aqadbtov Sttyaziga) ; Xenophon, An/i5. 2-4.8; 2.5.40; 3.4.13; 3.5.17; 4.3.4; Demosthenes, Orat 14,31; Ctesias, Pers, 20 (’Agzdjcavog), 23,29,30,31; Trogus Pomp. 10; Diodorus 15.2.1; 15,8,3; 15.9.3; 15.11.2; 15.90.3; 15.91,1; Polyaenus 7.14; Plutarch, Artaxerxes 27. Coins: B. V. Head, Hisloria numorum (2nd ed., Oxford 1911) 507; cf. D 428 nn, 4-5, 429 n. 8, 608 n. 7; JM 27, 10; Justi, JV«men5iicft 235 (No. 6); J. Miller, in RE 18/1 1164-1166 (No, 6); also Tam, From Xerxes to Alexander 20-21, — Sa¬ trap of Armenia (401 B. C.), led the Satraps^ Revolt against Artaxerxcs !! of c. 366-360 B. C., received, on submission, the Satrapy of Mysia. He married c. 401 B, G,, Rliodogune, daughter of the Great King Artaxerxes II, and died c. 344 B. C. Orontes II. Inscr. 393 (610) = 18 (34; cf. infra: Mithranes); cf. D 610 nn, 6-8^ for his filia¬ tion. — He must be the Satrap of Armenia of his name who led the Armenian auxiliaries of Darius HI in the battle of Gaugamela (1 October 331); Arrian, A nab, 3.8.5. It is extremely difficult to suppose that it was rather Orontes III (cf. Grousset, jffisfoire 79-80; H.Berve, in RE 18/1 1166 [No. 7]), because the latter most probably died after as late as 270/260B.C., i.e., some sixty or seventy years after Guagamela, and must, consequently, be presumed to have been a cente¬ narian at the time of his death, which is unlikely, Orontes II, on the other hand, could be in his middle sixties in 331. At the same time, it is equally difficult to accept — what JM propose (11 n. 2, p. 34) as a possibility — that Orontes II (HI according to JM, since the father of Artasyras is counted as the First; 10) reigned again after Mithranes down to about 316: for we 280 THE OHC»TTIDS OF ARIdENfA should then have to h^eve him to have died aged between 84 and 1401 Cf. infra Orontes III. I MiTHRAjtsfES (MithreneSj Mitfkhnes), Jnscr, 393 (610) — 18 (54: ^aatXea ... \ANhN rdv ^offtXSojg "‘A^odvdov); Diodorus 17, 21,7; 17.64.6; Arrian, AnaK 1.17.3; 3.16.5; Cuitius 3,12.7; 5.1.44; Dio Chrys, 73-2. The mutilated name of the above inscription, D 610 n, 6 would read, together with some others, as Oqddrr^v or Btt^ddvrjv; Honig- mann, 981^982, emends it as Mtd^dvfjv. This JM 34 do not accept. Honig- mann*s identification, however, accords well with the cumulative historical evidence (§7.) Also Justi, Namenbuch 214 (No, 1); Grousset, Histoire 79;Berve, in RE 15/2 2156. A coin has been attributed to him by O, Blau in WNZ 9 (1S77) too, but erroneously: E. Babdon, Les Rois de Syrie^ d*Arminie ei de Comma- gene (Paris 1890) cxci. — Iranian governor of Sardis, he passed to Alexander in 344 E.C,, and, in 331* was named by him to be Satrap of Armenia. ? Orontes III Diodorus 19.25.3 (c. 317 B.C.: "OptJvTot) rou if^ cf. Justi, Namenbuch 173- 174 {No. 4); Grousset, Hisioire 80. It is difficult to escape the impression that Xerxes was a son of Arsames, reigning as he did so soon after him and THE OBONTIDS OF ARMENIA 28$ being in possession of the city the other had founded. His father’s tribute that Xerxes iras obliged to renew may well have been imposed on Arsames by Seleucus 11 in retaliation for the support given to Antiochus Hierax. This filiation of Xersesj however, has not occurred to any specialists in the field, who usually regard him as not an Orontid,^’ the reason for this apparently being that Xerxes is not mentioned among the ancestors of Antiochus of Com- magene (infra § 5), But, as has been already noted (§2), not all the inscriptions have reached us in the Himrud-dag monument, and that monument itself has to do only with that king’s direct ancestors, whereas Ptolomaeus, the founder of the House of Commagene, could readily have been not a son, but let us say a nephew, of Xerxes. The genealogical position of Abdlssares is wholly unknown. The only reason for Supposing him to have belonged to the Orontid Dynasty and been, conse¬ quently, a King of Armenia, is the striking similarity of bis coins to those of Xerxes and, what is probably more significant, the identical shape of the head- gear of the two monarchs; cf. Visconti, Icon, greca 328-332, Ph xvi 4; Babelon Rois de Sgrie cxiv, 2! 1-212, PL xxix 3-5; Y. Langlois, Numismatique generate dt VAvm^nie (Paris 1659) 6-10; Head, Hist. num. 635; — (^uctMoii; (rd^ow); cf. Justi, Namenbuch 1; Baumgartner, in RE 1/1 26-27. Though Babelon suggests (cxciv, cxcv) that he may have been the father of Xerxes, it it is more likely, in view of what has already been said about the probable parentage of Xerxes, that he was a brother of his, 4, The last Orontid sovereign of Armenia was Orontes (IV): Strabo 11.14.15; cf. E. Diehl, in RE 18/1 1166 (N* 9), He and his brother Mithras (Mi0pd[r?7l5 ?), High Priest of the temple of the Sun and of the Moon at Arma- vira, are mentioned in the Greek inscriptions discovered there in 1927, as In¬ terpreted by A. I. Boltunova and Manandyan. One inscription contains an address of Mithras to King Orontes; another alludes to the King’s tragic death.“ Not unnaturally, the Armenian historical tradition, which Pseudo-Moses of Chorene preserved in his Hisiorg of Armenia, confirms and amplifies these data. There are several references. The most important, in Book 2, cap, 37-46, con¬ tains the account of the reign of King Eruand — of his brother the High Priest Eruaz (2.40, 48) — and of his struggle with, and overthrow by, King Cf. D 012, n, ft: qni manifeste non eiusdem afqaeArsamei domusesset; JM11 a. 4,29 (that the headgear of Xerxes and AtMlissares is dihereut to that of the King of Commagene, as Is remai'ked by JM, has nothing to do with their belonging or not belongiug to the Oroutida of Armenia). ^ Boltunova, 'OreCeskie uadplsl Armavlra,' TAFAN (1942) Nos 1-2; Manandyan, Arjna- viri hunaren arjeutagr.; ci. O iorgople 36-37. Treyer, Oc. po isi. Arm- 134-137 (lns is patent from the existence of a similar imbroglio in the Primary Histo¬ ry of Armenia^ where mention is made of a King Artashes, son of Sanatruces, while Eruand and his brother, also Artashes, are placed some two centuries earlier.®® This erroneous tradition found its way into Iberian historiography as well; and Leontius of Ruisi speaks of the Armenian kings larvand and his brother Artashan, as living in the first century of our era.®* The other references in Pseudo-Moses to the Orontids are merely passing: names inserted in the theogony of the early kings with which Book 1 of his work is concerned. Thus, Eruand is the father and predecessor of Tigranes, See infra ti. 27. 2,37 (157). ^ tbld^ For Sia^truces, &eo Aadourianj ArntM.^om 100-103; Debeyolse, Parfhia 235. MQuandyaa, O torgoole 3S-39. — Orontes, however. Is not made a son of Abganis. ®® AsdonrUn, 103; Debevolse, Parthia 217~21S, Soc ibid. 213-'2B0 for the in¬ volvement of lK>tli Armenia and Osrhoene In the Flomau-Parthlart politics In the age ot the Antonines. ^ Mauandyan, O torgfovle 48-53. Prim. Hist. Arm- 15-16. Leont. Mrov. 44-50. 's THE OHONTIDS OF AKhlHNEA 285 the ally of Cyrus the Great (1.22) j Aravan and Hrant are found respectively before and after him in the same theogony (1.31*19). The first of these referen¬ ces is of considerable interest. Artaxias, who overthrew the Orontids, was, to all appearances, a local dynast. The Greek form of his name renders, it has already been noted, AHa^es which is the purely Armenian form of the name, of which the Iranian form, ArfaxSaOra^ is usually Graecized as Artaxerseg.^^ The dynasty he founded, moreover, showed a marked preference for the name of Tigraues, four out of the eight Artaxiad kings bearing it. This was the name* it will be recalled, used by the proto-Armenian kings remembered by both Xenophon and the Armenian historical tradition.® "Whether Artaxias was descended from the proto-Armenian Tigranids* as Professor Adontz held to be indubitable,® or not, some genealogical connexion between them, real or pre- tented* seems definitely indicated. Now, the establishment of satrapal con- ' trol in Armenia, of which the Orontids were the beneficiaries, must have re¬ duced the rights of what local dynasts there were. Thus the struggle of Arta- xias and Orontes IV assumes the double character of a national revolt against an Iranian dynasty combined with the dynastic hostility of the Tigranids* or their relatives, for the satrapal Orontids. At all events, the above reference to the succession of Tigranes after Orontes — and it is the habit of Caucasian historiography to designate the facts of succession among remote, chiefly fo¬ reign, rulers by terms expressive of genealogical relationship® — allegorizes this same event as the succession of the eponym of the Tigranids after the eponym of the Orontids. 5. There are two assumptions that have always been made by historians in connexion with this period of Armenian history, which are quite gratuitous and incorrect, which have introduced much confusion into the little information we possess regarding the Orontids, and which must be obviated before a dear picture of the Orontid period can be obtained. One of these assumptions is that the south-western Armenian province of Sophene was a distinct State* separate from the rest of Greater Armenia, before the time of Artaxias and Zariadris®! (what the situation was before the Achaemenian phase, is not our HQbwliTiuiim* GraiianaUk 2S-29, 505; a different opinion, wbidi HDb^chmaDD does not accept, see Juati, Nanifn&ucA Se-S7, cf. 34-36- ® Supra at n. 7. ® Adontz* Armenija 390 n. 1, cf. 427 n. t. Ho even caJls the ArtaTdad? 'Tigranids* 389- 390. ® Cf. Ill/1 n. 40. The early kings remembecd in snch a confused way by the Armenian historical tradition were more remote than any foreign monarchs to the minds of those who gave it the literary form In which it has reached us. Cf., e.g., Manandyan, O torgotfU 30-40J Gronsset* Histoire 80; Hoiugmann, Kommagenc 980-981. 286 ■rae ORONTios of Armenia concern here). Thus, Manaiidyan“ speaks of the Orontid masters of Greater Armenia — and of the great trade route that passed through it connecting Iran and the Euxine and was controlled from the Orontid cities of Armavira and Eruandashat — as being distinct from the rulers of Sophene — a land closely bound to the Seleucid empire both culturally and economically. Ac¬ cordingly, he reckons Arsames, the builder of Arsamosata in Sophene^ among the latter and is unaware of his connexion with the former. Yet the evidence of the Nimrud-dag monument for his belonging to the Orontid Dynasty is incontrovertible. That one and the same ruling house should have held both Armenia's central plain and its Syro-Mesopotamian frontier, presents nothing extraordinary, unless the separation of Sophene and Greater Armenia be assum¬ ed a priori to have existed at that time. Actually, Manandyan himself sug¬ gested that the Orontids must have led the proto-Armenian expansion from the valley of the Arsanias — the land of Hayasa and, inevitably, also the land of neighbouring Sophene®® — to the valley of the Araxes, where they established their residence at Armavira.** What ground, then, can there be for conjecturing that they ever abandoned the southern part of the land which they had ori¬ ginally controlled? In this context, the following five points ought to be con¬ sidered. First, there is no Indication whatsoever that Sophene was politically distinct from Greater Armenia before Artaxias and Zariadris. Second, Strabo (11.14,15) is emphatic in stating the contrary : ^tarelxov t^v 'AqfiEvlav IHq- ffat ttai Maxsd6ve^t juerd zavra ol zifv ZvQiav l^^forre? xai M7}d(av * ZEXevzaio^ d' vyr^Q^ev ^Ogdvzi^g dbtoyovo^ "^Yddgvov, rcov iy^zd HsQtfcov er<^C ■ eZ6l’ z^v ^Avndxov zov peydXov azgarrjyfov rov yrgd^ TioXtfjL^oavzo^ dt^gdQri d[%at 'AQxa^iov xai Z/xQiddgto^ * xai o§roi, zov ^aoiXdox; ETiizQSipavtoQ * ^zrjB^vzog d* ixelvovt sigoaOepevoi *Pcofiaiotg xaB’ ai5zovg irdrzopto, ^a [iu the MSS ; jfiui^.uipattP\* L jfiuiq.tapuita uiin [in the MSS; II L. npipji^^i ptuuptiij .“« Markwart, ErapSatir 178; U | 7-8. ■■ Faustus 3.12. ^ Cod, TheodOB. 12.L3.6: satrapae Sophtuienae. Leont. Mrov. 53, 54. Ibid.; Roi^, List / 50 = Majr and Bri^re, Langue 671 023, conalders the name to bave been ' Dami ^ and ga.~ an enclitic particle;). ^ For the ArcxTini domains, ace 11 £ 12.;!. ^ Er^abr 177, 178. ^ Adont£, Atmenija 38, and Garitto, Hoaijnenta 234, consider the adjective in question as derived from a nomen gentiUcinmi 'Sahunl.' This name some ’would trace back to the patronymie Ma^uhit borne by □ Hittite prince (or perhaps dynasty) ruling over precisely western Sophene (5upn) and Mellteue In the eighth century B.C.: W. Bdek Soitrdge zar alteii Gwgrciphie Hnii Gesehicftle Vorderasicns I (Leipzig 1901) 50^52; Lap'anc'yan, 150 n. 1: Manandyan, O mk. apor. probL 132; II S 7 at n. 65. En 208 the Princes of Ingilene-Anzitene, Greater Sophene, and Lesser Sophene passed under the aegis of the Roman Empire; after 377, two more princely States from across the Arsantas shared their fate: II | 7. a* Ibid, £ 9. IbiiL n. 130. THE OBOa^lDS OF ARMENIA 305 Sennacherib; ’whereas Arzanene was, as has just been noted, the Arabian and not the ‘Assyrian’ March, the reference to Assyria being, obviously, due to the family’s connexion with the Orontid Kingdom of Sophene, which indeed had become the ‘Assyrian’ March. (5) The Vitaxae of Adiabcne, rulers of the Median March, were believed by Markwart to have been Orontid, and more particularly Artsrunid, and to have been transplanted by Tigranes the Great from Sophene, which was always ready to rise in revolt, to the Median frontier of his realm.^^^ — (6) TThe Princes Yahe’vnni or Yah(n)uni were the descendants of the divine Vahagn, the Armenian Heracles, according to Pseu¬ do-Moses (1*31), and also his hereditary high priests.^® As Yahagn was a grand¬ son of Emand, father Aravan, and ancestor of Zereh and Armog (supra § 11), the Orontid origin of this dynasty seems obviously indicated. — (7) The little- known dynasties of Aravenian and Zarehavan or Zarehavanian were descen¬ ded, according to Pseudo-Moses (1.31; cf. 2.8), from Aravan and from Zareh respectively (sEipra § 11), and are, therefore, to be regarded as Orontids; the latter honse may have been a branch of the Artsrunis.^® — (S) The equally little known Eruanduni dynasty (literally: Orontids), mentioned by Lazarus (chap, 70) must likewise be of Orontid origin.^ — (9) Last, but not least, there is the qnestion of the Orontid origin of the princely house of the BagratldSs^^ which will be considered in Part II. Ibid. 5 6. Ibid 'i Ibid. 5 12.7; £ 13.£3. Only one Ara’wene^tt la mentioned in the hjjstoricfd part ol Ps. Mosea's "wotk; (3.43), and none of the other hoii$e, \rhlch la mentioned only in the genealo^e^ of the early dynaete, at its beginning a separate branch. The mediaeval Ar¬ menian geograpbers distlngnlsh bet’ween the land at Zarehawan and that of Zarewand/Zara- wand, north of It and always coapled with that of Her; HUhschmann, Orfsnamcn 333. )@oth names are derived from Zareh; Markwart, ErdnSahr 177; Siidamteaieit 555-556. Ail the three lands lay in the Immediate vicinity of the chief AiraininiU domain of Greater Albak. In ’flew of etymological identity and geographical adjacency, the distinction between Zareha- wan and Zarewand appears somewhat artificial. The former was an Arcinnld land; Erdniala-f loe.cU, And in the lists of the AnuenlaQ Princes found in the doenments of the Gregorian Cycle (IX § 5) the Prince of Zarewand and of Her is next to him of the Aremnids, with but one other intercalated between them. The Prince oi Zarewand and of Her may, thus, have been the head of an Arcrunid branch. This would explain the absence ITom the sources of the mention of any members of this family: if it was not a separate family, its members were In¬ deed all Arcrunis. Ps. Moses, accordingly, preserved two separate versions regarding its origin; one for the Arcrunis, the other for the branch to which he applied the patronymic — rather than any territorial epithet — of Zarehnawan/ZerehawaneaQ; both, of course, pointing to the Orontid origin. He 'Was whoHy unconscious of this plurality and introdticed (1.6) still another one by etymologizing the toponym Zarewand; cf. Markwart, ^iuhznnenien 205 n. 1 (205-207), 555-556. ™ II 5 12,12. Ibid. £ 12.S. n 1* Regarding the Orontid origin of the Bagratid Dynasty, the Armenian historical tradition is quite emphatic. The Primarif Hisiorif oj Armenia}- iden¬ tifies the Orontid tutelary deity Anglf-Tork*') ■with the eponym of the Bagra- tids,® That the Bagarat-Angl of that monument is indeed that eponym, is evident from his being made the father and grandfather, respectively, of two other Bagratid eponyms, Biura't and Aspat,® A little farther in the same work (14), mention is made of the ‘great feudatory' Bagarat P‘‘ac[nav]a&ian, a descendant of Aramaniak, It has already been seen (I § 11) that the Orontids, as indeed a great number of Armenian dynasts, were deduced by the Armenian historical tradition from Aiamaniak, son of the divine primogenitor of the Armenians, Hayk, The prmnomen of that particular descendant of Aramaniak can leave no doubt as to his belonging to ‘the sons of Bagarat' On his part, Pseudo-Moses (1.22) also refers to the tradition of the Bagratid descent from Hayk; but he does this reluctantly and merely in order to reject it, with some vehemence, for the Hebrew origin that he propounds. What may appear to constitute an inner contradiction in this otherwise consistent tradition is that the Primary History represents Bagarat-Angl as a son of P'arnavaz, who is none other than the eponym of the Iberian royal house of the Pharnabazids (P^'amavaziani),* and that he applies* consequently, the Iberian royal surname, in its Armenian form of P'arnavazian, to Bagarat-AngTs descendant or kins¬ man just mentioned. This apparent contradiction is, however, patient of an explanation. But before this is attempted, Mark wart's thesis about the Pri¬ mary History and also Pseudo-Moses, and about the tradition they embody, must be examined. * Supra Part I | 11. * Ibid. § 14, lor the Orontid connexion witJi Angf. * This is overlooked by Abejy^n, 1st. dremearm^ Ilf. I 26, who fallB to note the relation of Bagatat to the Bagratids. Cf. also infra at n- 67, AspaL Aspet, for which see infra | 14. ^ That this P^afnawaz was indeed the eponym of the Iberian Pharnabazids Is clear from the reference (9-10) to his suhmisglon to Nahuchodonosor. The story of the Babylonian king’s connexion with the Iberians (of both Georgia and S^pain) goee back to Megasthenes (Seleucus I's ambassador to India) and has been preserved by Abydenus (in turn, preserved id Eusebius, Chron. I. 10 and Praepar. emjng. 9.41); by Josephus (in Con. Apiott. 1); and by Strabo (15.1.G, where, however, Sesostns is mentioned in this context instead of Nabuebodo- nosor). For this matter, sec Mark wart C^eal of this Study, this evidence was examined in connexion with Oron¬ tid history, and Indeed Orontid genealogy* and its relevance to both was established. Thus, of the names considered by Markwart as territorial sym¬ bols^ two are those of actual Orontid rulers: Xerxes and Zariadris.^ But the memory of the Orontids was also preserved in another way — precisely through their imprint on Armenian toponymy. Accordingly, the city of Bagaran in Ayrarat was built by the last Orontes^ and that of Eruandashat, in the same pTOvince, was the last Orontid capital in Armenia.*® It was doubtless also owing to the Orontids that there was a castle of Angl in Tsalkotn.*® Markwart him- ^f has once been led to infer from the evidence of the toponyms that some of the last Orontids — or the House of Sopheue, as he thought of them — were transferred by the Artaxiads to the south-eastern frontieT of Armenia.®* It is this migration that must account for the existence in those regions of the dynasties of Artsmni, Eruanduni, and Zarehavanian, which have all the ear¬ marks of being Orontids. And it is to these Orontid branches that the topo¬ nyms of those regions, derived from Orontid praenomina — E^ualldnnik^ Sha* varshan, Zarehavan (canton), and Zarevand — must owe their existence.® ^ It to be stressed again that the genealogical tradition In question is significant only as an embodiment of historical memories regarding personages and dynasties of the remote past (and dynasties, indeed, represented by eponyms), hut that no indication of the exact kinship binding these persons can be expected from it: cf, supra I’art 1, end oi § 9. Ps. Moses 2,40. Ct. g 11. Ps. Muses 2.39; cf. Manandyan, O torgovli 38-40. ® Lazarus 31 (127); Eiiseus 3 (74); SebSos 22 (102); Procopius Bell.pers. 2.25.5-15 (Vdyys^- For Cajkotn, see infra at n, 05. ** Supra Part 1 § le, at n. 117. ® Supra Part 1116. ^^ As has been seen ibid. n. 118, 'Zarebawanean* seems to have been Ps. Moses’s appellation ior the Orontid Princes of Her and Zarewand or Zarehawan. Oveiv looking the evidenoe for their existence, Markwart considers the above dynastic name a fig- ment of Ps. Moses’s imagination and a symbol for the city Zerehawan: Ceneatogie 20. It must be remarked, however, that in the Armenian princely nomenclature territorial names were derived horn cantons that constituted princedoms and not from towns. The cumula¬ tive evidence on the subject leaves little doubt as to the Orontid origin of the princedom in question. As for the origin of the name of the city of Zarehawan in Cajkotn, it Is true that Zariadris of Sophene could not have been its founder: Ceii6alogte20. It Is equally true, that THE ORONTIDS OF ABMENIA 311 In other 'words> it h by Oroutid memories thot both the toponymy and the genealogical tradition are explicable; it is, surely, hardly satisfactory to attempt to explain the genealogical tradition by the toponyms and leave the toponyms themselves unexplained. 5. Next, Mark wart attempted to establish the exact epoch and dr cum stances of the compilation of the Primary History. His thesis, as has been mentioned, is that it was produced at the Court and for the political ends of Bagarat, Piince of Taraun and Presiding Prince {‘Prince of Princes’) of Armenia for the Caliph, in the years 830-851,®^ This he set out to prove in a very complex argument. The eponym of the Bagratid house in the Primary History, Bagarat, is, as has been seen, identified with the god Ang| (I § 11). It has also been seen that Angl was the pagan Armenian equivalent of Nergal and that one of the manifestations of this equivalence consists in the fact that in the Armenian Bible, in 4 Kings 17.30 — in the phrase ‘and the Cuthites made NergeP — the name of the Cuthite divinity ts rendered by angel. Thb Markwart would regard as fortuitous, the Armenian word in question being, according to his intimation, an approximation to the Septuagint rendering of ‘Nergel’ as rajif No doubt, had we but this biblical text for the existence of that equivalence, Markwart's suggestion might appear convincing. How¬ ever, the cumulaive evidence cited in Part I § 14-15 tends to show, on the contrary, that, so far from creating, this text merely reflects the identity of Angl and Nergal. Precisely the same is the case of Hayk whose name was an Orontid Prince of the Zarchnwonld dynasty would bavo had nothing to do in Ca]kota. And yet that city, and another one named ZarlMt, appear to have been situated not far from Emanda^t which was indeed founded by an Orontid; cf. Manandyan, O t^rgosit 1X9, X20. Are we to suppose that the name Zareh appeared in the Orontid family before their loss of Armenia? What complicates the situation is that the Artaxiads, too, appear to have used it. The father of Artaadas I was named Zareh (Part I n. 59), and a son ol Tlgranes the (jreat appears to have been called Sarlaster (= Zariadris?): Valerius Masimns, 9.11 ext. 3 (cf. Appian, Mithr. 104); of. G^nealogie 21. For this personage, see Grousset, H Is to fre 349-351, 353-354, 358-359, 308-309; Laurent, Arminie 105-100, 117-118, 122-123, 120-127; cf. Markwart, S&darmenwn 390-298, 495. — Following the death in 826 of Aiot IV, Flince of the Bagratids and Presiding Prince of Arme¬ nia for the Caliph, the Bagratid domimons were divided chiefly between his sons Bagarat and Smhat, the former becoming Prince of Tarawa, Xoyt', and Sasun, and the latter. Prince of Siracene and ArSarunik^. The Caliphate, pursuing the pokey of dlttide el impera recognised Smhat as High Constable of Armenia and Bagarat as Presiding Prince ol Armenia with the title of Prince of Princes, In 83Q. The Caliphate was eminently suocessfal in its policy, and the relations hetw'een the two brothers and the branches they headed were not cordial, in 851, Bagarat fell from favour and was deported to Sdmarra, which he was not destined to leave. Geraaiflsie 51-53; cf, supm Part 1 114 at n. 92. 312 THE ORQNTIDS OF ARMENIA made to correspond to that of OrionMoreover, the phonetic proximity of ergel and angel would hardly have been patent enough to suggest the use of the one for the other, unless the essential affinity between the two religious ideas represented by these terms had already been familiar to the translator.®^ 6, Markwart’s argument continues as follows? the biblical name Cuthites, in Armenian, must have become confused with Xuf*; that of the inha¬ bitants of Khoyt', a canton of the province of Turuberan.®’ Thomas Artsruni, writing in the tenth century refers to the latter as "marauders ’ {meknakazenk'n Xiit*ay) and testifies to their loyalty to their prince, who happened to be Bagarat of Taraun.®® Accordingly, while admitting that he was not certain as to whether the pagan cult of Tork'" (whose identity with Angl [I § 14] he re¬ cognized) had persisted from pre-Christian days among the people of Khoyt^ Markwart nevertheless felt certain that when the author of the Primary Mis¬ ery wrote that "Bagarat was also called Angel and was in those days called a god hy the barbarous people’ the reference was not to the legendary eponym of the Bagratid dynasty, but to Bagarat of Taraun, deified by his subjects of Khoyt"' and, in the maimer of the pagans of Lystra thinking St Paul and St Barnabas to be Mercury and Jupiter (Acts 14.10-12), identified by them with their old pagan deity.®® There is, unfortunately, no evidence that the reference to a ‘barbarous people’ is anything other than the typical way in which many recently-Christian chroniclers were wont to refer to their recently-pagan an- cestors^ — and this may be an additional indication of the antiquity of the Primary History; or, finally, that any inhabitants, no matter how marauding of a land that had been Christian for some six centuries could possibly have manifested a completely polytheistic psychology that would have been per¬ fectly natural to pagans who had never known Christianity. I n. 1^8 The history of the Armenian biblical translatioas Is not vciy dear, but the first trans¬ lation of the Bible appears to have been made from the Syriac rather than from the Greek: Ahelyan Jai. drevnearm. tit, 1 S4-S5; Lyonnet, Origines. ® For this canton, see Htlbschmann, Orisnatntn 325, Thomas 2.7 (201) describes the mountaineers of Xoyt' In connexion with their revolt against the Caliph's forces after the deportation of Bagarat of Tarawn (cf. snpm n. 33); cf. Grousset, Histoire SSS-SSS. Geneatogie 52-54. ^ Cf-, e.g., among the Georgians, Leont. Mrov. 17; or the Russ. Prim. ChroR, 14-15; or in¬ deed Ps. Moses (snp«E at n, 17). The latter’s attitude, since he wrote at a greater distance from the pagan days (cf. infra $ 17), is more detached and more academic: he enhemedses and attributes unpleasant traits to some gods that bad been reduced to heroes, and he as¬ cribes the raising of the Idol of Vahagn to the Iberians (1.31), while admitting that he was an Armenian dlviiiity, but he does not tax his ancestors with being barbarians. THE OKONTIDS OF ARMENIA 313 7. Nextj MarkTfvart shifted somewhat his ground. Having just attempted to prove that Bagarat the Eponym had been modelled on Bagarat of Taraun, he went on the demonstrate that the latter had served as a model for the Eponym* s descendant, Bagarat P'arnavazian. Before examining this point, the text of the Primc(ry Hisiory (14) dealing with the latter personage must he quoted: firu^JupiaatiT t/"~ %iu^utptap ^utUtphtp^ t ^t/iu t^uitnatpat^ aai^^ A fi. ^ ub^p II ^ i^w^uiu*- t^ta ji ^'anp it ^ i^h-pmf nu^^tuupiutn ta^u^at^nt-tt. II ii/ii/ f*t.p f* ^1% 1 ^Qum tapmp atp^pay taua^h^ta ^tiytaautath irpi^p^^y wtp ^p L ^pt*t^ tfit^uiurntp p-inipaitJipnt.p-h^i^^t ^piadtubuiy It. ^ityp tup^pat^^y jAl t^i^payp* npiruP t-ia m^pnt^ uyiiitp^^ i ^*** ff^nrt»*iptritiy q^rf^nyr^My np tpvLprmt-prprt^ frij^yh jt i^pay ^apnt ^ J^jnuptity t There came to meet him“ with an army Bagarat P^amavazian, of the sons of Aramaniak, a great feudatory. He brought to him offerings of gold and silver and vested him with the robe and the ephod;^ he crowned him with the ancestral crown and seated him upon the throne of gold and precious stones; and he gave him his daughter in marriage. Him Arsaces the King made Aspet of the land of Armenia, that is, a prince and commander of the whole kingdom — the summit of authority, and [also] the King’s father and brother. And to him he gave the princedom of the realm. He crushed the giants who had raised levies against him in Mesopotamia of the Syrians. The name of Bagarat, the epoch that appears to be that of the early Arta- xiads,^ and the reference to Syria can leave but little doubt that the passage in question contains a memory of Bagadates, Tigranes the Great’s viceroy of ^ The legendary hrst Arsacid King of Armenia, Who was according to the Prim. Hist, Ann, 13, Arsaces, son of Arsaees the Great of Parthia (c. 250-24$ B.G.), or, according to Ps. Moses 2.3, Vologases, brother of the same; cf. supni Part 1 at n. 75. — For the pro|ectlo!i by the Artnetiian hUtorical tradition of Arsacid rule to the third century B.C,, see X at nti. 174-176. ^ 'Kobe' is a tentative readerlng of the fvord otherwise unknown; cf. Markwart, Genealogie t6n.l. For the roy^, no less than priestly, significance of the ephod (zniJIcos) lathe ancient East Mediterranean world, see M. Thiersch, Epetidgion und Ephod, Collesbild und Prlesierkleid im Alien Vorderasien (Stuttgart 19SQ}; Fraiue, Rogaat4 i&rail. 204-205. ^ Cf. stipTu n. 41. 314 TH& ORONTID^ QF ABMFS4IA Syria, mth whom we shall deal presently (§ 12), This Markwart recognised, hut* refusing to admit — wrongly, as we have seen — that memories so andent could have lingered ou, thought that this hit of information must have reached the compiler of the Primary History through some Greek or Syriac source." The reference to the crowning of the King by Bagarat and the marriage to him of the latter’s daughter, Markwart, in the same vein, would attribute to the influence of the text of Faustus, 5.44, telling how the Mamikonld Manuel, the all-powerful Regent of Armenia, set up C- 37$ the young Arsaees III as King of Armenia, gave him his daughter in marriage* and married the King’s brother Vologasea to the daughter the Bagratid prince* hereditary Coronant of the Arsacid kings.^ Once again, we seie the tendency to make of te:!ttual influence a substitute for actual history. The compiler of the Primary History, who, for all the ancient traditions his work may contain, could not have set it to writing prior to the invention of the Armenian alphabet on the threshold of the fifth century* was undoubtedly no less aware of the royal alliance and the hereditary office {for which, see § 15) of the Bagratids than was Faustus, and so could project, independently of the latter* both the office — which was natural* since so many Bagratids had by then already held rt — and the alliance to the epoch of their illustrious early ancestor. 8. But Markwart’s argument centred on something else. He would see, to repeat, in the figure of Bagarat P'^arnavazian, for all the admitted echoes of Bagadates, a projection into the past of the figure of Bagarat of Taraun. There are several points to this argument. First, it takes up the above text’s state¬ ment that King Arsaces created Bagarat ‘the King’s father and brother/ Accordingly, this is interpreted as inspired by another text of Faustus* in Book 4.14, where the historian alludes to some domains in Taraun that belonged — in the fourth century — to the Grand Chamberlain of Armenia. The chief appanage of that dignitary, however, was the Principality of Mardpetakan, in Vaspurakan, whence came his title of mardpct.^ His other title was indeed ‘the King’s father’ (hayr t'^agawori), so that he was usually referred to as hayr- mardpet So, Bagarat P'amavazian — we are told — was called ‘the King’s father* because he was modelled on Bagarat of Taraun* and Taraun it was where — five centuries earlier — the Grand Chamberlain had held some vil¬ lages/^ This is hardly serious. There is* moreover* a difference between ‘ the King’s father’ and 'the King’s father and brother’: the latter was not the title ** GencaioffU 56. " Ibid. 4B. " II 5 7‘S, Genealagie 48. In this connexion Markwart rather bums the man of straw In declaring absurd the Idea that the Bagratids ever held the office of Grand Chaniherlain. TEiE ORONTiDS OF ARMENIA 315 borne by the Grand Chamberlains. More than that, the former title was not seldom used in. the ancient East Mediterranean world to designate the chief ministers of kings: suffice it to recall here Aman in Esther 13.6; 16.11. The argument continues with the suggestion that the words of our text; ‘prince and commander of the whole kingdom" {(ixan ew hramanaiar amenayn t‘agaworut*eyna3ty. The first king of the new house was Mirvan I, related to the earlier kings in the female bne.^ His prae- nomen, a cognate of the Iranian Mihran, may suggest his belonging to the Iranian family of Mihran,®*® which, at a later date, indeed gave a royal dynasty to Iberia, that of the Chosroids,®^ The Orontids, however, also used this prae- jRamen, as in the case of Mithxanes ( == Mihran), son of Orontes II, and of Mithras (Mithranes?), brother of Orontes IV (I § 2). Moreover, the early Orontids could well have been described as Iranians. Finally, whereas we may merely infer the existence of the Mihranids at that early epoch, that of the Orontids is an historical fact*, and the moment of Mirvan Ts accession falls in the period of the Orontid Monarchy in neighbouring Armenia (I § 10). We have, in addition, the pattern of the Armenian royal cadets on the Iberian throne; thus, the rule of the Second Pharnabazid Dynasty was momentarily interrupted by that of a branch of the Artaxiads, then, at a somewhat later date, the Armenian Arsa- cids ruled in Iberia, and, finally, the Bagratids.® In fact the struggle of the Second Pharnabazid Dynasty and the Artaxiads in Iberia seems to have the character of a repercussion of the Orontid-Artaxiad struggle in Armenia. Ac¬ cordingly, if the Second Pharnabazid Dynasty was indeed an Orontid branch, Markwart, Giitealogie 74, wdulij interpret the name Armog, which designates. In the genealogy of the Prtm^ Hi»t. Arnt. and of Ps. the Orontid Artanes^Artoantes {stxpra. Part I at u. 72 and | 11), a& representing, instead, the King of Iberia, Artoees-Artag (in the Hr&t oentury B,G.). This change from 't' to *m' in the Prim* HisU (and, ioUowing it, in Ps. Moses) is the same as In 'Bagaram' and ‘Blvrtatn’ for Bagarat and *Blwrat.' €f. infra n. €0a, — The use of the patronymic derived from Biwrat by Asojik and Samuel of Ani in apphoaUon to Smbat VI (t 726/7) and Alot II (t ^90) Bagratnni Is an obvious archaeologlstn evoked by the memory of the above-mentioned second-century Stnbat; in the same vein, 'Vardan (76-77) oomparea to the same Smbat ASot FV Bagratani. ^ I at n. lOl and n. lOa, ^ The immemoTial antiquity of the YwpJihran houses (Ehtdcham, Iran Achim. 21 n. 4) may Justify this suggestion. On the other hand, at this early epoch, we know of no historical¬ ly ascertainable Mihranids. I n. 165; II | 25.J. ® For the Artaxiads of Iberia, see I n. 103; tor the Arsacids; ibiti* n. 105; and for the Ba- gratids: supra n. 53. 318 THE OKONTIDS OF ATRMENJA the indu^on of its eponym and its surname in the material containing memo¬ ries o! the Orontids as quite explicable.*®* 11. At all events, the cumulative evidence at our disposal quite forcefully indicates the Orontid origin of the Bagratids. We may now sum up this evi¬ dence. (I) Both the Primary History and Pseudo-Moses indicate this origin (this has already been noted in § 1); the latter indeed covertly, as a descent from the national and once-divine primogenitor Hayk, from whom Bagarat- Angh the founder of the Bagratids according to the Pnmctry History was him¬ self descended, and he mentions it only in order to refute it. (II) However, the vehemence of Pseudo-Moses's refutation is in itself an additional proof. It has been seen (I § 15) that the Christian scruples of the Orontid Houses of Arzanene, Artsruni and Gnuni forced them to exchange the traditional version of their descent from the Orontid tutelary deity Ang|-Tork' for a new one deducing them from King Sennacherib of Assyria. It was un¬ doubtedly the same sentiments that must have made the Bagratids abandon, in their turn, the Orontid claim and adopt instead a more general one that traced them to Hayk. The latter, it will be recalled (cf. 1 § 15), when reduced from his original position of part an astral deity and part a divinized primo¬ genitor to that of a mere hero, proved far less objectionable to the early Chris¬ tian writers than was the god of the netherworld and fertility, AngJ-Torl^:^ Nevertheless, the earlier, pagan, claim must still have remained patent —■ the descent from Hayk including that from Bagarat-Angl — in the new one; hence Pseudo-Moses, who calmly recorded the Haykid origin—when it implied no connexion with Angl — of other princely houses,® rejected it In the case of the Bagratids; and he proceeded to formulate an entirely different version that could match the new version of the Houses of Arzanene, Aitsuni, and Gnuni. To this we shall return shortly (§ 16). (III) A number of geographical and toponymical data point in the same direction, A link seems to be indicated by these data to have existed between the Orontid Dynasty and the nameBagarat and in particular one of its com¬ ponents;, the Iranian root baga^ indicative of divinity. The name itself, of course^ is an Armenian rendering of the Iranian *bagadata ('god-given’).®* In reference to n. 57 supra. It is to be noted that Artoce^ of Ibona was as Artaxiud, whose branch succeeded, and was followed by, what appears to be the Ibexian hranch of the Oron- tfds. It woidd have been perfectly natural, however, for the later Armenian tradition to have remembered among the latter. This might explain how —- under the Influence of that memory— 'Artoantes' could have been corrupted as 'Armog/Artog.' ® Cf. I at nn. 10S-17O and a. 173. Adontz, Aroienila 412; Markwart, EranSahr 174; Hbbschmann, Qronimafik 31. HDhsch- maim gives another possible Iranian etymology; *itagaMta {*god^a giftO^ The Xranold character of the name of the Bagratid eponym and of the Bagratid gentilltfal title (g 14) THE OHONTIDS OF ABMENIA 31d Now the Orontid kings of Armenia controlled, as is known (I § 5), the central Armenian plain — the valley of the Araxes. There, on that river, stood the last capital of the Orontids in Armenia, Eruandashat {— ^Orontasata); close by, the last Orontes oti hcaSligcooaj (In Q the soitassVW of A is replaced by the soireErif J' (region) of the south'] of the later M. P. 203: the Emperor came to Sper! in order to enter Iberia (fiiScqg.'OHji 1>5ig^iSQ, P, 226: the Byzantines seized Iberia's borders: Speri and the end of Cholarzene da®3g6t»a tija 3cas;;cq TSie first and the last passages are admittedly ambiguous, but the second one Is unequlyocal. !h this light, the border In question seems to have passed, precisely, between Sper and the 'end of Cholaxzene The referenceB to Speri (238) and to the ‘ Speri river' (=» Acampsis) and the ' sea of Speri' (=> the Euxine) found elsewhere in JuanSer and in Leontius of Ruisi prove nothing one way or the other. The neict reference, just as equivocal. Is in theTtwelfth-century Hist. Dasid JXI. Finally, the decisive fact is that, whatever the presumed ethnic connexion between the Saspeires mtd the proto-C^rgians, tiie Bagratid princes of Sper were within the Armenian political and coltural sphere. ^ Adontz, Armenila. 306-307; HDbschmann, Orfsiiomcn 364-365, cf. 441; Markwait, Getis- alogis li-12; Streilsage 252; Houignmnn, Osigreme 147 {Koao^iT of the Byzantines), In Faustus, 3.7; 4.40; 5,1, Dadwnk' (also: Darawnlt*, Darewnk', Daroynk% Darunk') appears as a royal fortress housing Axsacld treasures. Tins made Adontz think that the canton Itself had been a royal domain: Armenija 307, The Implication ot this was that the Bagratids could have acquired it only after the end of the Arsgcld MonEirchy. But occupying fortresses on THE ORONTID3 OF ARMENIA 323 or Tmorik% in Gordyeno;’® and possibly the canton of Colthene or Golt^n, in Siunia.” The havoc wrought by the struggle accompanying the establishment of the suzerainty of the Caliphate over Armenia {A.D. 653/4) led to a considera¬ ble change in the politico-dynastic configuration of the realm. Numerous dynasties came to extinction or grew weak to the profit of others; numerous allods changed hands.®*^ As a result, the Eagratids lost practically all of their original princedoms and acquired new ones; but the period of caliphal control is somewhat beyond the scope of this Study.® princfily terrltgry -was one of tlie royal prerogatives: 1 at n. 195; ct* the case of JngUeue: IX at nn. 66-70. ThiSi therefore, need not be regarded as equivalent to the royal possefsion of the territory itself. It Is not known when Kogovlt became a BagratidpTincedoni;possibly it had always been that, as a remnant of the Orontid appanage of 'Bagrat's region^ (cf, 111 [HI]); bat it was in the seventh century that Dariwnk'’ became the chief residence (osfon) and sepulchre of the dynaaty. There is. hardly any need to suggest with Adontz, 307, that, because a fifth-century Bagratid prince, Tlroc' I, took part in the affairs of the Great King's portion of Armenia (whereas Syspiritis lay in the Emperor's section), he must have belonged to a line different to that which held Syspiritis: a 'Persarmenian’ line which, as Adontz con¬ cedes, may already then have held Kogovit. The simple fact seems to be rather that the Bagratids as a house held simultaneously domains in different parte of Armenia; *cette divi¬ sion de leurs domaines italt... une cause de faiblesse; elle leor donnait par contre une grande security contre les tentaUves des maitres Strangers de TArm^nle. Qnand Us ^talent d’accord avec Byzance, ou quand Us redoutaient ieur voisin oriental, ila s^Journaient dans la priuci- paut^ de Sper. Mala Us a valent 5 Dariounk et dans son territoire, une forteresae et une princj- paut^, qui devenaient leur centre d'actJon quand ils fuyaient les Grecs, on quand Els Malent particuli^rcmenl en faveur auprfea des maftres de TArmenie Oricntalc'; Laurent, S6. ™ Tiimumi of the Assyrians, Tct/noQiTig of the Greeks (which some editors of Strabo, 11,14 5, have attempted to 'emend' as TagcavlTig wa$ also known as Kordrik*, but its chief fortress was always Tmorik*': Adontz, Armeni/a 395; Hisi. d’Arm. 210; Hubschmann, Orts- neanett 336-337; Markwart, SUdarmenien 350, 352-354, 383-386; cf. Garitte, Documents 2l9- 220. According to Ps. Moses, 2.53, the Bagratids were established in this princedom already in the secO'nd century* in the person of Smbat, son of Biwrat (for whom, see supra | 10). By the beginning of the eighth century, the Bagratids appear entitled ‘ Prince of the region of Vaspurakan' {Smbat iSzann kolmctim Vaspurakani}: Leontius 3 (27). Since practicatly the whole of Vaspurakan was held by various other dynasties (cf. Adontz, Armenija 315-321), this title must have been based on the Bagratid sovereignty over Its two limitrophe re^ons, Kogovit and Tamoritis; cf. Laurent, Arm^nie 80. ^ Ps. Moses 2.53 also mentions the setting up of a Bagratid foothold In Colthene. ff Ba¬ gratid indeed* the Princes of Golthene formed a separate branch, which Ps. Moses elsewhere (2.8) describes — owing, obviously, io the geograpbicol position of Golthene ^— as a branch of the House of Siunia: cf. II 3 12.IP. Cf. Grousset, Histaire 296-340; Laurent, S3-128. After 750, when the Umayyads, whom they supported, were overthrown by the Abba- sids (Grousset, HlsioUt 317-321), and especially after 772, when the Armenian revolt against the Caliphate, in which they took part, met with a crushing defeat (ibid. 323-334), the Bagra- tids suffered a temporary reversal of their fortunes and lost some of their possessions, Vas- purakan passed to the House of Arcnini. That Syspiritis was lost likewise, has been too 324 THE OliONTroS OF AfiMBNIA Now* if indeed the Bagratids originally held Bagravandene, it is not difficult to see why they came to lose it. The Ariaxiads are known to have transferred some Orontid branches from their original allods in Sophene to the Median border.^ A ^oriioTi then, they must have found the existence of an Orontid allod in the centre of Armenia itself quite intolerable. They may, we presume, have induced the Bagratids to exchange that allod for a more remote one, such as Syspiritis on the Iberian border. If Bagadates was the one who accepted this transactiou, the application to him in the PTumT^ History of the surname of Pharnabazid, indicative as it is of an Iberian connexion, becomes explicable. 14. Besides the surname of Bagratuni, i.e., Eagratid, the dynasty had anoth¬ er, short-lived, one of Aspetuni, which appears to have been the earlier of the two and was derived from the gentilitial title of aspti^ This title was derived by readily beliewd by Laumit, Anaiiiiie t09. The inaiaUattoa there by the Empeivr af a Ba- gratld prince In {tST need not be taken signifying the reiurn of Syspidtis to the dynasty. Actually, after 772, ASot IV took refuge In the Bagratld lands on the Impodal frontier, where he had stiver mines. This co«i1d only be Syspiritis. Already Steaho, 11.14.9, mentioned the gold mines of Syspiritis; and In the Ottoman kasa of Ispir, which represented a part at least of the old principality, there were gold mines, which were abandoned in the sisteeiofth een- silver mines, still used at the end of the nineteenth: Culnet, Turguie 1 160; ef. Laurent 41. This source of wealth enabled A^ot to bny from the Kamsarahans the prince¬ doms of Ar^arunik' and of Slracene. He then acquired also ASoc' and a part of Tayk'. He came, finally, we do not known how (Laurent 104 states that he wrested it from the Arabs), in possession o! MamUconid Tarawu. His grandson, Aiot the Great seized Mauiikonld Bagra- vandene. The Mamikonid princedom of Bznuulk' (with Xiat^) seems to have also been acquired by the Bagratids already by 760, though it was soon to become an Arab emirate. See, for all this, II | Markwart, Str^tfsiige 452; Genca/oyie 31; Laurent 93-98, 217 ; Grousset, 341, 373-374; T at n. 93. Il.aurent 93 states that the Bagratids acquired from the Mamikonids, e. 750, MnS andBaleS as weQ as Tarawn; but they appear luBa^atid hands only in the ninth century; and there is nothing in Leontius, 23, 33,34, quoted by Laurent, 93 u. 7, 99,94 and 110 n. 9, to support the assertion that Isaac Bagratuni was Prince of Tarawu, that Vasak Bagratuni was one also, and, a forttarU that Vasak expelled the line of Isaac from Tarawu (for these princes, see infra ill % 3). The most that the text of Leontius 34 warrants one to suppose is that Vasak held lands In the nelglibouihood of Arce^ and of Vaspurakan and so, according to Markwart, Stf&lfziigs 414-415, 'likely in Tamwn.' Some authors have attri¬ buted to the Bagratids the possession of fngiieiie, which Is a natural enough confusion arising from their Orontid origin: cf. Gronsset 292; Laurent 35 (where the Bagratids are, wrongly, given the title of miardpeU for which, see II g 7-8). Cf. infra Ell at nn. 26-2S. Supra at nn. 31-32, The Gk Life of St Gregory 9S: t < 5 v *0(S3ictbOiV&v iatdQxrfq; Procopins BelLp«rs, 2.3.12- 18:Tw'Ao7Tst‘£aW(>v... yivo^ — a phrase nusiaterpreted as ‘the tribe called the Asperianl' in the LCL ed., 1 [1914], 273; yet a little below (239/281), the same word y^vog is correctly rendered as 'family' [of the Arfiadds}. See Markwart, Streifzilffe 437; Adontz, Arateniia 402, 417. The fom * Aspetuni or, possibly, *Aspeietm Is not found In any Armenian sources, though It must have cndsted. THE OBONTIDS OF ARMENIA 325 Adonta from the Old Pers. vidapaitiS^ and, more convincingly, by Marfcwart and Hiibsclimann from the Old Pers. *aspapaiti^ or Master of the Horse,®® That this term designated an Armenian office, i.e., that of commander of the cavalry, as has been assumed by some;,*® I am now prepared to doubt. Adontz has made it quite clear®^ that there was no room in the Kingdom of Armenia for such an office, next to that of High Constable, because the Ar¬ menian army, which to all intents and purposes was exclusively cavalry, was under the authority of the latter; this, to my naind, is decisive. It must be asstuned, therefore, that, whatever the etymological significance of the ,term, it must have been merely a family title of the BagratidSi and not an office.®® There are indeed hardly any references in the works of the Arsacid historians to Bagratids in command of the king's forces.®® Wliat makes the whole question somewhat involved is the fact that one of the Seven Great Houses of the Iranian empire was sumamed Aspahbadh,* And this nomen genUHcium appears to have been derived not from any term signifying Master of the Horse, but, paralleUy with the Iranian term for High Constable, or spahbad, from the Old Persian spMapaitiL'^ May it not be sup¬ posed, then, that in a sitnilar way the Armenian gentUitial title aspet was derived, along with the name of the office of High Constable: sparapet or asparapeU Adontz op. ciL 401-402. Markwart, Gcnealogie 6S; Hub^chmann, Graininatik 109. d. Grousset, Histoire 291. Armenija 447. ®® For Annenlaii gentllitial titles, see Adoutz 400: a^p^t of the Gagratids, mamak of the Maxnlkontd^, Tnalxas of tbe XorKOTUois. The Prim. Eisf. Arm. 14, Indeed appears to imply that the King of Armenia created Bagarat P^arua^azean an aspel (| 7), bnt this must he due to the frequent fusion of this tiltle with the offiee of Corosast in the Bagraild nomen¬ clature; cf. supra n. €8. In Sa;3sanlaii Iran, on tbe other hand, the office of Gonunauder-in-Chief or High Con¬ stable = EFSR^spdhiaS (Christensen, Iran Sass. 180-132; Eht^cham, Iron Achim. 63-64: from the Old Pers. spSdapaittS; also Hhbschmann, Granvnatik 240) seems to have co-esisted with that of Master of the Horse = aspabad (Christensen 107-108, and n. 1; G. Huart and L. DeJaporte, L’Irtm antique [Paris 1952] 365) or {Ghjistensea 108 n. 1: this form 'seraltplus vralsembiahle'; it is derived from the Old Pers. *ospapaitiM or * 0 i 8 abilra- pailtl: Bht^cham 65); cf. Adontz, Armentja 447; Huart 365; Christensen 107-108, The chief source for the esiatence of the second office is Theophylactus 3.8; though Christenseti, 130-132, omits all meutioii of it when treating of the orgatilEation of the Iranian army. Christensen, Iran Sass* 103-105; Eht^cham, Itan Ack^m^ 21 n, 4; Justl, Nantenboich 306, 429. ^ Ghristeaseo, op. off. t04 n.l, Ps. Moses, 2.68, derives the name of the Iranian iamlly (A^pahapeti Pahtaw) from their position as commanders of the armed forces. 'What is especi¬ ally Interesting, Theophanes, Citron. 352, refers to the Aspahhad of the end of the fifth cen¬ tury as 326 THE CRONTIDS OF ARMENIA from the same Old Persian $pMap however, the some¬ what baffling existence not only of parallel offices, in Armenia and in Iran, which was to be expected in the circumstances, but also of the parallelism between two great houses, one Iranian and the other Armenian, bearing etymo¬ logically equivalent names of Aspahbad and of Aspet. To be sure, there were other such paralleHsms between the two societies. The Armenian Kamsarakans and Gregoiids claimed to be branches of the Iranian houses of Karin and Su¬ ren.^ After the Hellenistic phase of Armenian history, stretching from the Orontids to the advent of the Arsacids in the first century, a new phase of * Ira- nianism' was entered by Armenian society: the impact of the Parthian empire of the Arsacid Dynasty of which the Armenian royal house was a branch. In this new phase, the Armenian aristocracy must have begun to pattern itself on the Iranian, exactly as the Arsacid Monarchy of Armenia tended to become institutionally a mirror of the Parthian empire. Thus it is entirely possible that, either through sheer imitation or for reason of a marital alliance, the Bagratids assumed injts Armenian form the appellation of the Iranian Aspah- hadhs. 15. The great hereditary office of the Bagratids was that of Goronant or fagadir of Armenia. The passage of the Primary History cited above (§ 7), which refers to the placing of the crown, along with the conferment of other regalia, by Bagaral P'ainavazian, and the documents of the Gregorian Cycled are the earliest references to this Bagratid office-fief. A similar office existed in Iran, where it w^as vested in the House of Suren,*® and may have served as a model for the Armenian office, though Adontz supposes something like the same office to have already existed in Urartu.*® The Bagratids appear to have held another office-fief, mentioned only in the Greek Life of St Gregory, of Guardian of the Caucasian and Tzannic mountains.*^ It implied the control of the Pontic Alps (Mt Paryadres), the north-western boundary of Syspiritis, and the position of a sort of assistant-vitaxa of the Nortli.®® The reference to Mt Caucasus, however, must be regarded as purely rhetorical. 16. With the Christian phase of Armenian history, succeeding the second 'Iranianism' of the Parthian phase, new fashions in genealogy were ushered into the princely society of Armenia. It has been noted (§ 11 [IIJ; I § 15) that, in the new phase, some Orontid branches abandoned their pagan tiadi- For sporapff, see HObscbmann, Grammatik 240,* Adoiitz, Armentja 445; cf. II J a. Ps. Moses 2.27, 28, 72, 73, 74, 82, 90, 91; II S 12./6; | 13.21. « II J 5 Lists A and B. Ghlistensen, Iran Sass^ 18; Eht^CtiAm, Iran AcMm. 21 n.4 23). “ Adontz, Hist. d’Arm. 215. II S 5 List A. ^ II 12,S. THE OHONTiDS OF ARMENIA 32 ? tion of the descent from the god Angl-Tork'. And Tvhile the Bagratids at first retained the vaguer, and less objectionahle, claim to a Hay kid descent, the Houses of Arzanene, Axtsruiii, and Gnuni devised a wholly new one, to the Assyrian royal ancestry. The vogue of exotic origins among the Armenian Princes® enhanced the basic, religious, motivation of the change. It was in these circumstances that the Bagratids in their turn evolved an entirely new genealogical tradition, of Hebrew origin. Pseudo-Moses appears to have been the formulator of this new theory. Now the new Arzanene-Artsmni-Gnuni claim was prompted by the conjunction of a geographical synonymy (‘City of Angr = Xity of Sennacherib') with a biblical tradition (the flight of the sons of Sennacherib to Armenia). In a somewhat similar way, it was the con¬ junction of several near-homophonies and of two historical traditions, Jewish and Armenian, that seems to have given birth to the Hebrew claim of the Bagratids. Markwart has suggested that Pseudo-^Moses must have been struck by a series of near-homophones in Josephus. There is a mention of Ananus, son of Bamadus {Bell. jud. 5.13.1) and of his companion (and one referred to together with him) Archelaus, son of Magadates (sttcf, 6,4.2) — contemporaries of Titus — as well as the story of the High Priest Ananelus, Herod the Great’s creature {Ant 15.2.7). These names, Pseudo-Moses must have correlated with the memories of Baga dates, Viceroy of Syria (g 12); and thus evolved (2.24) a composite and imaginary personage, theBagratid ‘Enanos the Aspet,’ whom he made journey to Palestine, at the time of Herod the Great, and take part in Jewish affairs, of which he was cognizant through, precisely, the works of Josephus. These seem to have been the steps that led to the idea of the Jewish origin of the Bagratids. When once formulated, this origin was emphatically asserted throughout the History of PseSdo-Moses (1.22; 2.3, 8,9, 33,63). In connexion with this, Pseudo-Moses was able to indulge his love for etymolo¬ gizing, when he proceeded (2.63; 1.22) to derive the typically Iranoid Bagratid praenomina of Bagarat, Smbat, Ashot, Varaz, from the Hebrew names Baga- dia (Bagath), Shambat or Shambay, Asud, Azaria or Vazaria, and, by impli¬ cation, the title of Aspet from the name Sap'atiay (Saphatias).^® At any rate, even if Markwart’s explanation be deemed unconvincing, the fact must never¬ theless be recognized that no other Armenian source prior to the tenth-century John the Katholikos knows anything about the Hebrew claim of the Bagratids. So, if not by Pseudo-Moses, this theory must have been developed at the time of Pseudo-Moses, and he at least must be credited with putting it in its earliest- known literary form. Later, this theory underwent an important change. It was transformed into a tradition of such magnificence as outshone the genea- Cf. I at ti. 245. MarL'wari, SrSn^hr 174 n, 6 <= 175); Sti^ifsiige 426-430. 328 THE OEONTIDS OF AEMENIA logical chim^res of other dynastie&r the imperia] Chinese tradition of the Mami- konids and the royal Assyrian tradition of the Artsrunis. The latest Bagratid claim was one of their descent from King and Prophet David of Israel, the an¬ cestor of Onr Lord, the descendant in an unbroken line from Adam, and the archetype of kings. This implied that the Bagratids were not only the most ancient and, as it were, the most authentic dynasty in the world, but, moreover, kinsmen of Our Lord and of His Mother. This version arose and developed, however, not among the Armenian Bagratids, but among their Iberian cousins. The Iberian (East Georgian) line of the dynasty stemmed from Atrnerseh, or in Georgian: Adamase, son of Yasak and grandson of Ashot III the Blind (t 761), Presiding Prince of Armenia. Following the defeat of the Aimenian insurrection against the Caliphate in 772, Adamase removed to Iberia thus founding this line.^“^ The Georgian sources^ at all events, are the tirst to men¬ tion the Davidic origin of the Bagratids. The earliest source is Juansher’s History of King Vokhtang GorgasaU written between c. 790 and c. SOO, where is related the arrival in Iberia, sometime after 772, of the above Adamase, ‘who was of the House of David the Prophet, The next source is the stone effigy in low relief of Adamase’s son, the Guropalate Ashot I the Great of Iberia (t 830)“® from the church of Opiza, in Shavshet'i, which represents him in an act of offering a model of that church to Our Lord, seated upon a throne, blessing Ashot, and accompanied by the King-Prophet, represented In an attitude of supplication and identifiable by the ecclesiastical majuscules CDYT' {Cinas- carmeigueli DuViT' — ‘the Prophet DavidHere the allusion to the don- or*s descent from Our Lord’s ancestor and the latter’s intercession is unmis¬ takable. The life of St Gregory of Khandzt'a, written in 950/951 by George Merch*ule,™^ is next to refer, in Chapter 11, to the tradition of the Davidic origin as extant at the time of the Guropalate Ashot the Great. Finally, the Chronicle of Iberiat compiled in the eleventh century, mentions this tradition as existing at the time of Ashot’s father Adamase.^®® From the latter source ^ Infra III; IV j 34-35; Bagr. of Iber. I. JeauSer 243; cf. IV | 34, and n. 26. The date his death is discussed in my Chivnologg 83-S5. Amh-^nasylU, Isi. graz^ te*. 1 212-213 and Table lit- ASot is identifiable by an inscrfp- ticn. In another inscription from the same church, he is called the 'second hirilder' of it: Mbit, Dneanik poje^iv^opSeiiju. i Ktardietifu {St Petershurg 19il> 163. He must have been so called because, according to Juau^er 178 the original builder of the monastery of Opiza ■was Artavaz, Duhe Of Qhdtarzene, femp^ Vastang I (late fifth centuiy)- In the CAron. Iber, 260, however, it Is Aiot's yeaugeet son Gnaram who is qualified as the second builder of Opiza: no doubt as the second, after bis father, among the Bagratids. Cf. Introd. at nn, 54^. «W CftJWi- Iber. 251; cf. IV J 34, and a. 27. TBE OKqwrriD^ QF AltMKNIA 329 and from Juanaher^ it appears that the dalm was not, in the days of Adamase, as yet widely l53iown.“" This would suggest that it had just then come into being. The first Armenian author to refer to this new theory — rather in passing — was John Katholikos (f 931), as he was the first after Pseudo-Moses to mention the Hebrew theory.^ Constantine Porphyrogenitus (f 959) was the first to do so among the Byzantines. The latter’s account betrays a definite Iberian inflnence and manifests an attempt at representing this theory in the form of a genealogy,™ This genealogical formulation reached its complete and elaborate shape in the Georgian work of Sumhat, the Hislory of ffte BagTaiids^ written about the year 1030.^ ™ Unis, the Choaroid princess nt Iberln, whose son married Adamase'^s daughter, is ahuwn by the CAron. 251, to have been ignoiaut of the Davidic origin of the Bagratids; and Juodier, 243, While saying, on the one hand, that Adamase was Of the House of David, on the other hand describes Adamnse's father as ‘related' (actueaiy ‘related by marriage* = mzaxebat [sic]) to the Bagratlde; cf. in/ra III at n. 21; IV | 34. ™ John Rath. 25; qapdt f^Bagarat.,, who is renowned for being of the House of David’). The Bagarat in question 1$ the same as the one referred to in the Peiat. Hist. Arm. as 'Bagarat P'arnawazean* (|7). In generai, the ArmeDian Bagratids displayed Jlttte interest in the Hebrew theory and its Davidic de¬ velopment, as compared with their Iberian eousins with whom this legend became the basis of a dyuastio-poUtlcal myth. One may say that Ps- Moses launched this Idea for the use of hla ArmenlBU patrons, but that it was their Iberian kinsmen who made use ol It. The connexion of this Idea with the Armenian Bagratids exists largely only in modem historiography, moulded as it has been by Ps. Moses: It is enough to remember the casual reference of John Kath. to realize this; cf. ht/ror n. 110. ™ adm. fizip. 45; cf. IV S 34 n. 35. The Iberian inspiration is evident in Coustantiue's avoidance of all mention of the Armenian past of the Iherlan Bagratids; this tendency reached its fullest expression in the History of Sumbat. It® IV Bxcursus A. — It is incredible that an attempt sbonld have been made to-day to rehabilitate the theory of the Hebrew and Davidic origin of the Bagratids. En JenkLns, ed.. Coast. Potphgr. Ds Adm. Imp. IE, Runrimah suggests (172) that ‘we need not doubt their [the Bagratids'] Jewish origin: large numbers of Jews from the Assyrian captivity... settled in Armenia, where, as In Babylonia, there were hereditary chiefs who claimed descent from David known as the "Princes of the Diaspora” till the high Middle Ages.* The relerenCe grven is to M. Brann and D. Chwolsou, ‘Evrei,* ES 11 (1S03) 44CMi41, where, as a matter Of fact, there is not one reference to Armenia, this entire section dealing with the Jews, and their BxUarchs (rsS gaiuta), in Babylonia. Referring to the same work, H. Rosenthal, 'Ar¬ menia,' JE 2 (1902) 117, introduces the words ‘adjoining Armenia,' when speaking ot the descendants of the Jewish captives of NebnchodonosOr 'in the Parthian and Persian coun¬ tries.* The presence of Jewish settlers In Armenia is, of course, generally known; and H is just aa weS known that the Exllarchs did not reside there. At all events, the entire attempt loses significance when It Is recalled that the Davidic claim originated not in Armenia, but in Iberia, and that even the Annenian theory ot the Hebrew origin is a late one; It cannot be found In any source prior to the eighth-centuiy work of Ps. Moses. The latter*s ascription of 330 THE ORONTIDJS OF ARMENIA VI. The correlation of the versionsj Armenian and Iberian, of the third, Jewish, genealogical theory of the Bagratids bears upon the problem of the date of Pseudo-Moses. For over half a century the problem of the trne date of the composition of his great work on the Armenian Antiquities has taxed the ingenuity of scholars. Pseudo-Moses himself supplies his readers with broad hints which make it possible to place his jhTuii in the second half of the fifth century^ and which were once accepted* as now they are not, at their face value by the overwhelming majority of specialists. Being, tlius, something of a de¬ liberate mystifier, this author has deserved his appellation of Pseudo-Moses, As to the true date, scholarly opinions vary. Broadly speaking, there are three groups of theories on this subject; some ascribe this History to the seventh century,^® others to the eighth,^® and still others to the ninth.These di¬ vergent views, and their mutual exclusion, were recently held up to irony by a Hebrew (not Davidic) origin to the House of Aiuatunl, which Runcluian assumes to be a cor¬ roborating evidence. Is made rather hesitantly, along with another and totally different claim, which Is probably the earlier one. For the possible origin of this claim in Ps. Moses, see II I £2.^. Somewhat more plausible, though as incapable of proof, Is the suggestion of Maoler that the Hebrew theory may have been due to the conversioti to Judaism of some pre-Christian Bagratids (as in the case of the House of Adiabeue): Risioire t7n/«efselte par Btienni Asallk de Tar6n (Paris 1917) 7 u, 9 (= ■&), In either case, the history of the claim is the best fliv gument against its validity. Ps. Moses $.61, 62, 6$; cf. Abelyan, fst. dreumarm. Itt, i 198-1&9, 207. ^ See, e.g., A. von Gutschmld, ' t)ber die GlaubwUrdlgkeit der armenischeu Gescbidhte des Moses you Khoren,' BVSGW 27 (1876); and his article on 'Moses of Chorene,' completed by F.C. Conybeare, £B llth ed. (19Ii) [between A.D. 634 and 643: this must he regarded as the definitive opinion of the last-named scholar]; A. Zaminean, grakaii patmuVisvA (Nakhiehevan 1914) 110; L. Mellkset-Bek, ‘Xazary po drevnearmjansMm istocnikam v svjazi sprobiemoj Moiseja Xorenskogo,’ Sf>ornik v iesi’ Akad. I.A. Ob^U 112-118 [cf. to/ra n, 120]. ^ E.g., Cairitrc, RoudcIIcs sourcet de Molse de Ktiaren.\ Etudes critiqae^ (Vienna 1893); JVoiioelfSs sDurcCs da Moise de Khoreni Sapplimint (Vienna 1894); G. XalOteaoc*, Armfan&kij Epos If 'Istorii Arntenii* Moiseja. Xorenskago (Moscow 1896); AtJttfatiskle ArSakidg v *Iatorii Arwfinii’ Moisefa Xoretiskago (Moscow 1903); G. Ter-Mkrtican, ‘Xprenac'woy £amanak§ oroSela nor p'orj,’ At 1897 [after the end of the seventh century]; H. Akiuean, ‘Moses Ghorena^],' 'RESuppl. 6 534-641 [Ps. Moses is identical with Leontius tbe Priest, c. 800]; Adontz, ' Sur la date de I'Histoire de I’Arminie de Moise de Chotone: k propos de 1‘article de M, Hans Lewy,’ B 11 (1936) 97-100; 'A propos de la note de M. Lewy sur Molse de Chorine,' itld, 597-599 [between the last quarter of the eighth ecntuiy and S26]; Jena^ia, K krliike 473-503 [not before the eighth CJentury], ii* E.g., Mlaker, ‘Zur Geschichte des Ps- Moses Chorenatsi,' Ar 2 (1927); 'Die Datlerang der Geschichte des Ps. Moses Xorenacl,' WZKM 42 (1935) 267-286; Markwart, Gmeaio&ie [mid-ninth century]; Manandyan, Xorsnac^ii af^lcoaoi tucumi (Erevan 1934) [second half of tbe ninth century]; H. Lewy, ‘The Date and Purpose of Moses of Ghoreue's History,' B 11 87-96; 'Additional Note on the Pate of Moses of Ghorene,' lifid. 503-596 (between 876 and 885]. HIE ORONTIDS OF ARMENIA 331 a Soviet-Armenian savant, wlio reverted to the traditional date;^^ and he has been seconded in this by another authority,^® While pointing out, quite justly the mutually cancelling divergencies of modern scholars when dealing with the question of when Pseudo-Moses did write, these two Soviet-Armenian authori¬ ties seem to overlook the fact of the solid agreement of all of them as to when Pseudo-Moses could not have been writing, i.e., in the fifth century. The best among the arguments against the traditional dating of Pseudo-Moses appear to me to be the following. <1) frr 1,14, Pseudo-Moses projects into a remote past the division of western Armenia and some neighbouring lands into First, Second, Third, and Fourth Armenia, which division was instituted by the Emperor Justinian in 536.“^ — (2) In 3.18, he speaks of the Iranians’ pene¬ trating as far as Bithynia in the course of a war on the Empire. This occurred, for the first time in history, in the war of 604^29.^^® — {$) In 3.46, allusion is made to the institution, following the death of Arsaces III (c. 390), of the office of Presiding Prince {arajiiord naxornrac'n), along with that of comes Armeniae {komess Uxam) in the provinces fallen under Imperial control. This can only be a reminiscence of the situation which resulted from Heraclius I’s victory over Iran in 629. — (4) In 2,65, he refers to the Khazars (as at the time of the mythical First Arsacid King Vologases), which no Armenian source does prior to the Geography of Ananias of Siracene, of the end of the seventh century, once erroneously ascribed fo the same Pseudo-Moses, At the beginning of that century, Sebeos does not mention the Khazars by name.^™ — (5) He makes use, in 2.62, of ‘Vaspurakan’ to designate the territory east of lake Van; this territory, however, came to be so designated only after the partition of Armenia in 591.^ Sebeos, in the early seventh century, does not yet know this term as a toponym, but uses eospufakan adjectivally as an ‘elevated’ equivalent of ‘ Iranian’ and thus indeed to designate the territory in question, which in 591 remained in the Iranian sphere.^® It is only in the Narratio ™ S. Mpisasyaac', Xorenac^u ofejicnacl Surfe (Elrevau 1940). ^ Abelyan, 1st. dJvmeoFm, Hf. i 19S-209. Adont!. Anneni/a 203; Gutschn^id-Conybeare, Mosr# itf Cfiorene S98 n.l. ^ lifid. ™ Eutrod. a. 11. ™ Melikset-Bek, Xasarg. The author^a Intention seems to bo Isas to provo Ps. Mosos's bo- longing to the seventh century than to dcmonstr&te that no Armenian source of unquestion¬ able daUng prior to Ananias knows the Khazars. He does not altogether exclude the possi¬ bility of an interpolation; but the presence of other anachronisms makes this possibility extremely unlikely. — Sebeos, who does not use the term ‘Khazar,' nevertheless refers to the Kbazar king as 'great Xak'an of the North': IS (104, 106), 19 (lOS, 109). Adontz, Armenifa 230-234. ^ SebSos 3 (40); m 0 (70, 77), t^aspuraican is as unmistakable synonym ot ‘ Iranian.' Cf. AdonU, AJ'iTtentja 232; Garitte, La N